We initiated a new field of yeast genetics with our discovery that two non-chromosomal genetic elements, [URE3] and [PSI], were prions (infectious proteins), analogous to the agent causing the transmissible spongiform encephalopathies of mammals. The first, [URE3], is an altered form of Ure2p, the protein product of the chromosomal URE2 gene important in regulation of nitrogen catabolism. The second, [PSI], is an altered form of Sup35p, a subunit of the translation release factor and product of the chromosomal SUP35 gene. We found that Ure2p is more resistant to protease digestion in [URE3] strains than in wild-type strains, and is aggregated specifically in cells carrying the prion, supporting the prion model and suggesting amyloid formation as its molecular basis. The N-terminal 65 aminoacid residues of Ure2p is sufficient to propagate [URE3], or to induce the de novo appearance of [URE3]. We showed that the Ure2p prion domain forms amyloid filaments in vitro. Moreover, just as the prion domain induces prion formation in vivo, it induces the otherwise stably soluble native Ure2p to form amyloid in vitro. The properties of Ure2p amyloid formation in vitro reflect and explain the prion properties of [URE3] in vivo. We thus proposed that the [URE3] prion is an infectious amyloidosis. We showed that fragments of Ure2p or fusions with other proteins cure the prion efficiently. This phenomenon may be due to interruption of the growth of the amyloid 'crystals' due to the fragments or fusion proteins, and suggests a new approach to the treatment of amyloid diseases. We find that the Mks1 protein is essential for the de novo formation of the [URE3] prion. Mks1 activity is negatively regulated by the Ras - cAMP pathway, and we find that activation of Ras2p prevents de novo [URE3] prion formation by inactivating Mks1. We showed that the Hsp104 chaperone is necessary for [URE3] prion propagation, and that overexpression of the Hsp40-family chaperone Ydj1p can cure the [URE3] prion. We also showed that the Hsp70-family chaperone Ssa2p is also necessary for the propagation of [URE3]. Cells with the [URE3] prion contain networks of filaments consisting of the Ure2 protein. Further, most of the Ure2p in extracts of [URE3] strains is in a form insoluble even after boiling in 3M urea and 2% SDS, confirming that it is in an amyloid state. Our collaborators, Drs. Tim Umland and David Davies (LMB, NIDDK), have determined the structure of the nitrogen regulation domain of Ure2p and find that it is closely similar to glutathione-S-transferases (GST). Ure2p is inactivated by prion (amyloid) formation in vivo. We find that Ure2p is not inactivated by a conformational change in the functional part of the molecule, but by a steric effect or diffusion limitation on the interaction of Ure2p with Gln3p. The Ure2-GFP fusion protein forms amyloid filaments with a helical form. The length of the helical repeat is constant within each filament, but this length varies by more than 2 fold from one filament to another. This may be the basis of prion strains, that have different infectious properties and different effects on the host. We have isolated homologs of the URE2 gene from other strains of S. cerevisiae, from various pathogenic Candida species and from a filamentous fungus. While the C-terminal domain is highly conserved and the homologs can substitute for the cerevisiae Ure2p, the N-terminal domain (up to residue 99) is highly variable. Nonetheless, there is a conserved part of the prion domain from residues 10 to 39. This region apparently interacts with the Ure2p C-terminus as judged by inactivation of Ure2p when the fragment is overexpressed. This region also is responsible for the curing of the [URE3] prion by fusions with GFP mentioned above. We find that the prion domain forms the central core of the amyloid filaments, with residues 1 to 65 comprising the highly protease resistant part. Ure2p residues 71-95 serve as a linker between the amyloid core and the peripherally arrayed functional domains (residues 95-354). Monomers of Ure2p are bound to eachother by interactions between the prion domains. We have recently described an entirely new class of prions, based not on amyloid formation, but on the requirement for autoactivation in trans of the vacuolar protease B (PrB) of yeast. Cells that lack active PrB remain in that state, except for the rare (10^-5) spontaneous activation of the enzyme. Cells with active PrB give rise to progeny nearly all of whom have active enzyme. These cells can also infect cells without active enzyme by the transfer of active PrB. Thus PrB, in its active form is an infectious protein (a prion). There are many proteins that are necessary for their own activation, which could thus potentially act as prions. The N-terminal prion domain of Ure2p (residues 1-90) is rich in N and Q residues and we showed that these are important for prion formation and propagation. However, we find that there are no essential amino acid sequence elements in the prion domain: five random shuffles of these amino acids leave a protein that can form prions in vivo and amyloid in vitro. This surprising result implies that amino acid composition, rather than sequence, is the critical driving force for prion formation in the case of Ure2p. Likewise, we fing that randomizing the Sup35 prion domain does not abrogate ability to form a prion. These results suggest that amyloid of Ure2p and Sup35p have a parallel in-register beta sheet structure. Amyloid of Ure2p has an amyloid core that is high in beta sheet structure with beta strands perpendicular to the filament axis as judged by X-ray diffraction and electron diffraction (with U. Baxa, D. R. Davies, A. C. Steven). This result confirms the amyloid nature of the filaments formed by Ure2p. We find that introduction of amyloid formed in vitro of recombinant Ure2p infects yeast cells with the [URE3] prion, confirming that self-propagating amyloid is the basis of [URE3]. Several [URE3] variants arise in such infection experiments, and extracts of each variant infects cells with the same variant. The infectious material of recombinant Ure2p or cell extracts is larger than 20 nm in diameter (or at least 40 Ure2p monomers). Amyloid of recombinant Ure2p is essentially as infectious as extracts of [URE3] cells. We surveyed 70 wild strains of yeast and find that none carry either the [URE3] prion or the [PSI+] prion, whereas several selfish nucleic acid replicons (RNA viruses, DNA plasmid) are found in varying proportions of strains. This implies that these yeast prions are a net disadvantage to their host. Efforts are underway to find other chromosomal genes affecting prion generation, and to find new prions of yeast or of other organisms.